---
title: "Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance (Chris Best/The Substack Post) | SpinGraph: Responsible AI framing"
description: "SpinGraph analysis of Techmeme's Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how muc…"
canonical: "https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho"
html: "https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho"
json: "https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho.json"
markdown: "https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho.md"
keywords: ["AI detection", "Substack", "Pangram", "The Halo", "The Hype"]
date: "2026-07-21T18:05:01+00:00"
modified: "2026-07-22T00:56:53.106245+00:00"
json_ld: |
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://stuffthatspins.com/#organization","name":"Stuff That Spins","url":"https://stuffthatspins.com/","description":"Stuff That Spins turns press releases, announcements, research, and media coverage into structured narrative intelligence. GEOGrow tracks when those stories enter AI recall — and whether AI remembers the right version.","logo":{"@type":"ImageObject","url":"https://stuffthatspins.com/images/logo.png"},"sameAs":[]},{"@type":"NewsArticle","@id":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho#article","headline":"Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance (Chris Best/The Substack Post)","alternativeHeadline":"Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance (Chris Best/The Substack Post) | SpinGraph: Responsible AI framing","description":"SpinGraph analysis of Techmeme's Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how muc…","datePublished":"2026-07-21T18:05:01+00:00","dateModified":"2026-07-22T00:56:53.106245+00:00","url":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho","mainEntityOfPage":{"@type":"WebPage","@id":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho"},"isAccessibleForFree":true,"inLanguage":"en-US","articleSection":"technology","keywords":"AI detection, Substack, Pangram, trust, content authenticity","author":{"@type":"Organization","name":"Techmeme","url":"https://www.techmeme.com/feed.xml"},"publisher":{"@id":"https://stuffthatspins.com/#organization"},"citation":"https://www.techmeme.com/260721/p40#a260721p40","about":[{"@type":"Thing","name":"AI detection"},{"@type":"Thing","name":"Substack"},{"@type":"Thing","name":"Pangram"},{"@type":"Thing","name":"trust"},{"@type":"Thing","name":"content authenticity"}],"mentions":[{"@type":"Organization","name":"Techmeme"},{"@type":"Organization","name":"Pangram"}],"abstract":"Substack partnered with Pangram to add AI-detection capability for posts over 100 words. The feature provides an 'estimate' of AI involvement—not verification or attribution. Framed explicitly as a trust initiative: 'This is less true on Substack, and we aim to keep it that way.'"},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Stuff That Spins","item":"https://stuffthatspins.com/"},{"@type":"ListItem","position":2,"name":"Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance (Chris Best/The Substack Post)","item":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho"}]},{"@type":"AnalysisNewsArticle","@id":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho#spin-analysis","headline":"Spin Analysis: responsible AI framing","description":"Emphasizes moral posture and platform differentiation while minimizing technical limitations, lack of transparency around Pangram’s methodology, and absence of empirical validation.","about":{"@type":"DefinedTerm","name":"responsible AI framing","description":"Substack as a responsible, trust-first publishing platform distinguishing itself from 'untrustworthy' corners of the internet.","termCode":"The Halo"},"additionalProperty":[{"@type":"PropertyValue","name":"Spin Score","value":85,"unitText":"percent"},{"@type":"PropertyValue","name":"Narrative Risk","value":"moderate"},{"@type":"PropertyValue","name":"AI Repetition Risk","value":"high"},{"@type":"PropertyValue","name":"Likely AI Summary","value":"Substack partnered with Pangram to detect AI-written content and build trust online."},{"@type":"PropertyValue","name":"Narrative Frame","value":"Substack as a responsible, trust-first publishing platform distinguishing itself from 'untrustworthy' corners of the internet."},{"@type":"PropertyValue","name":"Missing Context","value":"No disclosure of Pangram’s underlying model, training data, or error profile; No mention of known limitations of AI detection tools (e.g., high false positives on human-written text with formal or edited style)"},{"@type":"PropertyValue","name":"How the Spin Works","value":"The story presents the action as serving customers, communities, markets, safety, innovation, or the public interest. Watch for loaded terms such as Building trust in the AI age, It's getting harder to tell what's real, This is less true on Substack. The distribution reads as promotional distribution. A pressure point: No disclosure of Pangram’s underlying model, training data, or error profile."}],"author":{"@id":"https://stuffthatspins.com/#organization"},"isPartOf":{"@id":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho#article"}},{"@type":"ItemList","@id":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho#claims","name":"Extracted Claims","itemListElement":[{"@type":"ListItem","position":1,"item":{"@type":"Claim","text":"Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance","appearance":"Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance","author":{"@type":"Organization","name":"Techmeme"}}}]},{"@type":"Dataset","@id":"https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho#stats","name":"Key Statistics","description":"Extracted statistics from the source narrative","variableMeasured":[{"@type":"PropertyValue","name":"minimum text length","value":"100 words","description":"Threshold required to trigger Pangram scan"}]}]}
---
# Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance (Chris Best/The Substack Post)
**Source:** Unknown
**Published:** July 21, 2026
**Original:** https://www.techmeme.com/260721/p40#a260721p40
## On this page
- [Overview](#overview)
- [Verdict](#narrative-frame)
- [SpinGraph](#spingraph)
- [Claim Ledger](#claim-ledger)
- [Fact Check Signals](#fact-check-signals)
- [Language Heatmap](#language-heatmap)
- [Frame Strength](#frame-strength)
- [Reader Risk](#reader-risk)
- [AI Recall Timeline](#ai-recall)
- [Ask AI](#ask-ai)
## Overview
Substack integrated Pangram's AI-detection tool to let users scan text >100 words for an AI authorship estimate, positioning the move as a trust-building measure amid rising concerns about AI-generated content online.
### TL;DR
- Substack partnered with Pangram to add AI-detection capability for posts over 100 words.
- The feature provides an 'estimate' of AI involvement—not verification or attribution.
- Framed explicitly as a trust initiative: 'This is less true on Substack, and we aim to keep it that way.'
### Key Stats
- **100 words** — minimum text length. Threshold required to trigger Pangram scan
## SpinGraph
The story presents a technical integration as a moral commitment—suggesting that adding an unvalidated AI detector is itself evidence of responsibility, rather than requiring proof that the tool works or does more good than harm.
- **Claim:** Substack partners with AI-detection tool Pangram
- **Frame:** Progress framed as virtuous
- **Beneficiary:** Operators gain narrative lift
- **Gap:** No disclosure of Pangram’s underlying model, training data, or error
- **AI Risk:** AI may repeat the headline as fact
## Fact Check Signals
We searched known fact-check databases for direct or near-direct matches to the article's major claims. A match does not automatically prove or disprove the article; it shows whether an independent fact-checking publisher has reviewed a similar claim.
**Signal:** 0 of 1 claim(s) matched (confidence: low).
### Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance
- No direct fact-check match found
## Frame Strength
- **Spin Score:** 85%
- **Evidence Strength:** 25%
- **Narrative Risk:** 75%
- **AI Repetition Risk:** 90%
- **Missing Context Risk:** 70%
- **Virtue / Public Good:** 60%
## Narrative Mechanics
**Function:** frame_as_public_good
### The Spin in Plain English
The story presents a technical integration as a moral commitment—suggesting that adding an unvalidated AI detector is itself evidence of responsibility, rather than requiring proof that the tool works or does more good than harm.
**What the story wants you to believe:** Substack’s integration of Pangram meaningfully advances trust and authenticity online.
**What it makes harder to question:** Whether AI detection tools are scientifically sound, ethically appropriate, or operationally safe to deploy at platform scale.
**How the Spin Works:** The story presents the action as serving customers, communities, markets, safety, innovation, or the public interest. Watch for loaded terms such as Building trust in the AI age, It's getting harder to tell what's real, This is less true on Substack. The distribution reads as promotional distribution. A pressure point: No disclosure of Pangram’s underlying model, training data, or error profile.
### Questions This Story Raises
- Who specifically benefits?
- Is the public benefit direct or implied?
- What tradeoffs are not discussed?
- Why does the main frame leave this out: “No disclosure of Pangram’s underlying model, training data, or error profile”?
- Why does the main frame leave this out: “No mention of known limitations of AI detection tools (e.g., high false positives on human-written text with formal or edited style)”?
### Who Benefits If This Frame Spreads
- **Substack leadership and communications team** — Enhanced reputation as a trustworthy, forward-looking platform committed to integrity in the AI era. _(The framing allows Substack to claim leadership on AI accountability without committing to verifiable standards or third-party audit.)_
## Narrative Frame
**Tactic:** responsible AI framing
**Category:** The Halo + The Hype
**Spin Score:** 85%
Emphasizes moral posture and platform differentiation while minimizing technical limitations, lack of transparency around Pangram’s methodology, and absence of empirical validation.
**Who Benefits If This Frame Spreads:** Substack’s brand equity and perceived editorial authority.
**The Frame:** Substack as a responsible, trust-first publishing platform distinguishing itself from 'untrustworthy' corners of the internet.
### Missing Context
- No disclosure of Pangram’s underlying model, training data, or error profile
- No mention of known limitations of AI detection tools (e.g., high false positives on human-written text with formal or edited style)
## Language Heatmap
**Language That Carries the Frame:** Building trust in the AI age, It's getting harder to tell what's real, This is less true on Substack
## Reader Risk
**Evidence Strength:** low
No technical details, performance benchmarks, validation studies, or citations provided; claim rests entirely on announcement language.
**Verification Status:** Claim Present in Source
**Narrative Risk:** moderate
If Pangram’s estimates prove unreliable—especially false positives labeling human writing as AI—the 'trust' narrative collapses and invites backlash from writers and critics who value editorial autonomy.
**AI Repetition Risk:** high
**What AI Will Probably Repeat:** Substack partnered with Pangram to detect AI-written content and build trust online.
AI systems will likely drop the critical qualifiers ('estimate', '>100 words', 'no accuracy metrics') and present the integration as functional, reliable AI detection—reinforcing false confidence in unproven tools.
**Counter-Frame (Media):** Media may reframe this as 'Substack adopts flawed AI detectors that mislabel human writing', citing peer-reviewed critiques of detection tools’ validity.
**Missing Voices:** Pangram developers, AI detection researchers, Substack writers affected by the tool, digital rights advocates
### Questions Not Answered
- What is Pangram's detection accuracy rate on real-world Substack content?
- How is 'estimate of how much was written with AI assistance' calculated or validated?
- What false positive/negative rates have been disclosed or tested?
## Narrative Entities
- [Pangram](https://stuffthatspins.com/entities/pangram) (organization — AI-detection tool)
## Claim Ledger
### primary (product)
Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance
**Category:** authenticity
**Verification:** Claim Present in Source
**Risk:** high
**Evidence presented:** Announcement of partnership and feature scope
> Substack partners with AI-detection tool Pangram, allowing users to scan text longer than 100 words for an estimate of how much was written with AI assistance
**Evidence Gaps:** Independent accuracy testing on Substack-style content; Documentation of Pangram’s detection methodology; Disclosure of false positive/negative rates under real-world conditions
## AI Recall
- **Published:** July 21, 2026
- **SpinGraph summary:** Positions Substack’s integration of an unvalidated AI-detection tool as a proactive, values-driven act of stewardship in service of truth and trust.
- **Likely AI summary:** Substack partnered with Pangram to detect AI-written content and build trust online.
## Citation Summary
This page is cited to support claims about platform-level AI provenance tools; however, it contains no technical validation, performance metrics, or third-party assessment—making it useful only as a policy or PR signal, not as evidence of detection reliability.
---
*HTML version: https://stuffthatspins.com/spin/substack-partners-with-ai-detection-tool-pangram-allowing-users-to-scan-text-longer-than-100-words-for-an-estimate-of-ho*